Machine Learning Readmission Risk Modeling: A Pediatric Case Study

Table 5

Average standard deviation f-score (F) performance [%] of SVM, MLP1, MLP2, and NB for decreasing number of folders in the RCV process. no SMOTE = no oversampling correction of class imbalance is done.

nfolds	SMOTE
	SVM	MLP2	MLP1	NB
10	9.85 0.63	7.54 0.16	10.23 0.8	9.83 0.3
5	9.67 0.49	7.65 0.19	9.67 0.26	9.83 0.53
4	9.6 0.17	7.65 0.02	9.71 0.23	9.83 0.13
3	9.73 0.18	7.57 0.06	9.69 0.07	9.98 0.17
	no SMOTE
	SVM	MLP2	MLP1	NB
10	1.86 0.00	9.70 1.45	NA	11.23 1.37
5	2.04 0.00	9.16 0.82	NA	11.20 1.14
4	1.95 0.40	9.62 0.75	NA	11.29 1.32
3	2.22 0.45	9.60 0.52	NA	11.08 1.23